Date of Completion

January 1982


Psychology, Social




This study sought to investigate the consequences of varying both the sources of group attraction and group size for group productivity. It was proposed that when the source of group attraction is collective attraction to the group task, cohesiveness will be positively related to performance. However, when intermember attraction is the source of group cohesion then cohesiveness will be negatively related to performance. A second hypothesis proposed that "social loafing" (i.e., decreases in member task effort corresponding to increases in group size) will occur when task attraction is low. Social loafing will not occur, however, when group task attraction is high. In such groups internal pressures to maximize one's effort counterbalance the tendency to slacken personal effort as group size increases. A final hypothesis predicted that as interpersonal attraction increased, group members were less likely to make egocentric attributions for group performance. As task attraction increased, however, members were more likely to make such attributions.^ One hundred, fifty eight subjects, working in either two or four person groups, were asked to construct as many paper objects (called "moon tents") as they could in a fixed time period. Task attraction was manipulated by stressing in high task groups the importance of the experiment and its results. Furthermore, these subjects were offered additional experimental credit if their group recorded the highest performance. Low task groups were not given either the importance information or the chance to earn extra credit. Interpersonal attraction was manipulated by having high interpersonal subjects interact for twenty minutes prior to the performance period. The interactions were structured so as to maximize conditions for member liking. Low interpersonal subjects did not interact, using the pre-performance period instead to complete a survey. During the performance period the frequency of member interactions were recorded to provide an index of group process loss.^ Two way analyses of variance performed on manipulation impact measures, while verifying the effectiveness of the manipulations, also revealed an unexpected main effect of interpersonal attraction on task commitment: As interpersonal attraction increased, a corresponding increase in task commitment also occurred. A three-way groups nested within treatment analysis of variance performed on performance scores revealed that, as expected, increasing group task attraction resulted in performance increments. Increasing interpersonal attraction, although resulting in an increase in interaction frequency, had no effect on performance. A path analysis procedure was used to test the post hoc explanation that increases in task commitment resulting from high member liking were mitigated by process loss effects revealed in the interaction frequency data. This analysis indicated that the effects of heightened task commitment accrued from high interpersonal attraction were indeed negated by the consequences of increased conversation resulting from this same attraction.^ The second hypothesis that task attraction moderated social loafing was fully supported. The third hypothesis, which proposed that both interpersonal and task attraction influenced the tendency of group members to make egocentric attributions, was supported only after accounting for the finding that no group perceived itself as failing.^ The potential implications of these results for both the cohesiveness/performance literature and the social loafing literature were discussed. Furthermore, the applicability of the data to organizational development interventions were also considered. ^